Model comparison
GPT-5 vs Qwen3.5 397B-A17B
GPT-5 is the stronger model overall, scoring 50.9 to 45.7 on the Noometry Index. Qwen3.5 397B-A17B costs 2.5× less per token, which makes it the better buy when GPT-5's lead doesn't matter for your workload.
Last verified . 29 shared benchmarks.
Alibaba (Qwen)Summary
- They share 29 benchmarks with published results for both. GPT-5 scores higher in 7 categories and Qwen3.5 397B-A17B in 3 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in long context, where GPT-5 leads 69.5 to 44.1.
- The biggest single-benchmark swing is FrontierMath (Tiers 1-3): 55.4% for GPT-5 and 31.2% for Qwen3.5 397B-A17B.
- Qwen3.5 397B-A17B is cheaper at $0.60 / $3.60 per million input/output tokens, against $1.25 / $10 for GPT-5.
- GPT-5 accepts more context: 400K tokens versus 262K.
- Qwen3.5 397B-A17B has downloadable open weights; the other is API-only.
Side by side
| GPT-5 | Qwen3.5 397B-A17B | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 50.9 | 45.7 |
| Released | 2025-08-07 | 2026-02-01 |
| Weights | Proprietary | Open |
| Context window | 400K | 262K |
| Max output | 128K | 66K |
| Input $ / M tokens | $1.25 | $0.60 |
| Output $ / M tokens | $10 | $3.60 |
| Results tracked | 69 | 36 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5 leads
GPT-5: 50.7 (#39), Qwen3.5 397B-A17B: 41.7 (#108)
| Benchmark | GPT-5 | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena WebDev | 1417 | 1400 |
| LMArena Coding | 1436 | 1464 |
| SWE-bench Verified | 73.6% | — |
| SWE-bench Verified (bash only) | 65% | — |
| Aider Polyglot | 88% | — |
| SciCode | 42.9% | — |
| GSO | 6.9% | — |
| WeirdML | 60.7% | — |
| ALE-Bench | 1,162 | — |
| AlgoTune | 1.67 | — |
Agentic & Tool Use Too close to call
GPT-5: 33.0 (#53), Qwen3.5 397B-A17B: 33.2 (#51)
| Benchmark | GPT-5 | Qwen3.5 397B-A17B |
|---|---|---|
| Terminal-Bench | 49.6% | — |
| APEX-Agents | — | 24.9% |
| GDPval | 34.8% | — |
| Remote Labor Index | 1.7% | — |
| τ²-bench Airline | — | 81.5% |
| τ²-bench Banking | — | 9.8% |
| τ²-bench Retail | — | 84.4% |
| τ²-bench Telecom | — | 97.8% |
| DeepResearch Bench | 49.6% | — |
| BALROG | 32.8% | — |
| LMArena Search | 1133 | — |
| METR Time Horizons | 69.6% | — |
Reasoning GPT-5 leads
GPT-5: 38.3 (#55), Qwen3.5 397B-A17B: 33.6 (#62)
| Benchmark | GPT-5 | Qwen3.5 397B-A17B |
|---|---|---|
| Kagi LLM Benchmark | 72.7% | 73.7% |
| Chess Puzzles | 37% | 13% |
| LMArena Hard Prompts | 1415 | 1448 |
| Mystery Game Puzzles | 23% | 18% |
| DTBench | 90.7% | 87.5% |
| LMCA | 40% | 37.9% |
| Epoch Capabilities Index | 150 | 146.64 |
| ARC-AGI-2 | 9.9% | — |
| SimpleBench | 56.7% | — |
| NYT Connections (extended) | — | 58.9% |
| ARC-AGI-1 | 65.7% | — |
| CritPt | 12.6% | — |
| EnigmaEval | 10.5% | — |
| Thematic Generalization | — | 65.1% |
| EBR-Bench | 12.7% | — |
| ForecastBench | 61.4 | — |
Math GPT-5 leads
GPT-5: 54.6 (#36), Qwen3.5 397B-A17B: 44.9 (#64)
| Benchmark | GPT-5 | Qwen3.5 397B-A17B |
|---|---|---|
| FrontierMath (Tiers 1-3) | 55.4% | 31.2% |
| OTIS Mock AIME 2024-2025 | 91.4% | 88.9% |
| LMArena Math | 1406 | 1455 |
| FrontierMath Tier 4 | 22% | — |
| ProofBench | 18% | — |
| Omni-MATH | 64.7% | — |
| MATH Level 5 | 98.1% | — |
| FrontierMath (Feb 2025 set) | 32.4% | — |
| FrontierMath Tier 4 (v1) | 12.5% | — |
Knowledge GPT-5 leads
GPT-5: 56.5 (#37), Qwen3.5 397B-A17B: 53.3 (#53)
| Benchmark | GPT-5 | Qwen3.5 397B-A17B |
|---|---|---|
| GPQA Diamond | 86.2% | 86.4% |
| LMArena Expert | 1420 | 1461 |
| Humanity's Last Exam | 25.3% | — |
| SimpleQA Verified | 50.1% | — |
| MMLU-Pro | 86.3% | — |
| Confabulations | 10.3% | — |
| Vectara Hallucination Rate | 14.7% | — |
| GPQA (HELM) | 79.2% | — |
Multimodal GPT-5 leads
GPT-5: 46.1 (#12), Qwen3.5 397B-A17B: 40.0 (#42)
| Benchmark | GPT-5 | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Vision | 1232 | 1262 |
| GeoBench | 81% | — |
| VPCT | 66% | — |
Multilingual Qwen3.5 397B-A17B leads
GPT-5: 51.5 (#105), Qwen3.5 397B-A17B: 54.0 (#56)
| Benchmark | GPT-5 | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Non-English | 1397 | 1431 |
| LMArena Chinese | 1423 | 1501 |
| LMArena French | 1409 | 1461 |
| LMArena German | 1415 | 1447 |
| LMArena Japanese | 1411 | 1427 |
| LMArena Korean | 1361 | 1382 |
| LMArena Russian | 1406 | 1429 |
| LMArena Spanish | 1398 | 1443 |
Instruction Following Qwen3.5 397B-A17B leads
GPT-5: 74.0 (#107), Qwen3.5 397B-A17B: 75.2 (#73)
| Benchmark | GPT-5 | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Instruction Following | 1388 | 1424 |
| IFEval | 87.5% | — |
Long Context GPT-5 leads
GPT-5: 69.5 (#2), Qwen3.5 397B-A17B: 44.1 (#71)
| Benchmark | GPT-5 | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Longer Query | 1399 | 1442 |
| Fiction.LiveBench | 97.2% | — |
Writing & Preference GPT-5 leads
GPT-5: 63.5 (#63), Qwen3.5 397B-A17B: 62.5 (#74)
| Benchmark | GPT-5 | Qwen3.5 397B-A17B |
|---|---|---|
| LMArena Text | 1406 | 1438 |
| LMArena Creative Writing | 1365 | 1401 |
| EQ-Bench Creative Writing | 1627 | 1478 |
| LMArena Multi-Turn | 1426 | 1447 |
| Short-Story Creative Writing | 86% | — |
| WildBench | 85.7% | — |
Frequently asked questions
Is GPT-5 better than Qwen3.5 397B-A17B?
GPT-5 is the stronger model overall, scoring 50.9 to 45.7 on the Noometry Index. Qwen3.5 397B-A17B costs 2.5× less per token, which makes it the better buy when GPT-5's lead doesn't matter for your workload.
Which is cheaper, GPT-5 or Qwen3.5 397B-A17B?
Qwen3.5 397B-A17B is cheaper. It lists at $0.60 per million input tokens and $3.60 per million output tokens; GPT-5 lists at $1.25 and $10.
Is GPT-5 or Qwen3.5 397B-A17B better for coding?
GPT-5 scores higher on coding benchmarks: 50.7 versus 41.7 in the Noometry coding category.
Which has the bigger context window?
GPT-5 does, with 400K tokens against 262K.
How many benchmarks do GPT-5 and Qwen3.5 397B-A17B share?
29 benchmarks have published results for both models. GPT-5 has 69 scored results on Noometry and Qwen3.5 397B-A17B has 36.